13 июня 2016 г.
Start talking about "artificial intelligence" and you'll get nowhere pretty fast. Most people are simply creeped out by the whole concept of thinking machines, conditioned from a young age to fear Hollywood's portrayals like HAL from 2001:A Space Odyssey and MCP from Tron. Ironically, nobody thinks of the helpful Rosie from The Jetsons or the adorably annoying C-3PO, who needs no other introduction. If they aren't creeped out, they'll probably just express disbelief: machines aren't really intelligent, and what current AI software does isn't really learning.
While there is some truth in this argument, the fact is that AI is coming an incredibly long way, and is doing so at a breathtaking pace. Significant improvements have been made in just the past couple of years, with more forthcoming constantly. Some of the world's most brilliant human minds are hard at work on building the smartest possible non-human minds, including Stephen Hawking, the think tanks behind Google and Facebook, and most of the major software vendors.
Though not a finished product, that doesn't mean that AI doesn't have some value to bring to the organization. There are many ways in which AI is already making improvements in business. No, this technology isn't cheap. Yes, it's getting better all the time. Where is there a place for AI in your organization?
Artificial Intelligence Can Reveal Hidden Patterns in Your Data
That's a lot of data. How will we ever analyze it all? Somebody call in the robot.
Any data sets that you would like to analyze for hidden correlations and patterns, AI technology is your guy. AWS cloud services utilizes this same technology to hone and refine their search algorithms. If your business is undertaking big data analytics or uses similar algorithms that you want to refine, AI is the best technology available. On the downside, the skills it takes to do this kind of analysis doesn't grow on trees.
Be prepared to search far and wide for the talent you need, or to partner with a firm that specializes in AI to get the skill and expertise you'll have to have. Unless you're willing to build a rather expensive team to handle your AI endeavors, Plan B is probably your best bet -- contract with an AI firm that already has its ducks lined up.
Artificial Intelligence Can Personalize Your Website Visitors' Experiences
If you've visited The North Face website multiple times, you've probably noticed their impressively personalized shopping experience. This experience was brought to you by a software company that specializes in AI technology. The website software literally "learns" what you like, look for, and do when visiting the website. Each time you go, it gets better at providing you with product recommendations. This type of customer personalization can also be applied outside the website experience.
For example, if your website "learns" that Customer A likes hiking boots and lives in Colorado, you can send them a heads-up via email when your Durango store runs a sale on hiking boots. Or, send Customer A an email coupon for 20 percent off her next pair. Better yet, send it for her birthday -- that kind of personalization is what will separate the businesses that excel versus those who hold Going Out of Business sales over the coming years.
Artificial Intelligence is Making Headway in Business Forecasting
What does the market look like next year? In five years? What products will be in, or go out? AI is good at these types of business forecasting, which require the analysis of far too many variables for humans to contend with.
Pan Am Airlines was a mainstay in air transport from World War II, until it wasn't anymore. It just disappeared. Blockbuster video was the place everybody went practically every week, until they didn't anymore. Now we all Netflix. Remember when everyone's dad drove an Oldsmobile? Until no one did, because they just weren't? The point is, these were thriving businesses -- literally household names and classic American icons -- that just dropped out of existence because the times changed and market demand shifted, leaving them in the dust.
What business wouldn't want the ability to avoid the fate of Pan Am, Blockbuster, and the Olds? AI is a powerful tool for business forecasting, which is quite difficult for humans to do, because it requires analysis of a large number of factors. Determining the future trends in an almost infinitely complex global marketplace requires a lot of data and a hefty load of analysis, but it isn't something you nail on the first try. AI software can learn over time and get increasingly better at business forecasting. It might be enough to foresee trends like the public skipping the weekly movie rental in lieu of some video streaming -- meaning you could gradually convert your vast movie collections to digital format and avoid the dreaded abandoned storefronts and gradually decaying signs that video stores left dotting the American cityscape.
Artificial Intelligence Can Handle Dynamic Pricing
Some businesses, such as Airbnb, offer dynamic pricing schemes, based on demand and availability in a given area at a given time. AI is quite good at this sort of thing. The algorithms are excellent at quickly gauging supply and demand and adjusting the pricing offered to customers based on predetermined ratios. Any number of businesses could benefit from this kind of dynamic pricing, such as those in the travel and hospitality industry, transportation, energy, fuel, and others. These pricing models could also work for various e-commerce businesses.
Beware of Artificial Intelligence for Customer Service
We've discussed a wide range of areas where AI can be very helpful to businesses, but that doesn't mean that machine learning is good for everything that it could potentially handle. Such is the case with customer service. While AI is actually quite good at determining what a given customer wants or needs and directing them to the right place to get help, customers hate it.
Part of the issue is that people want to talk to people, not machines. Another problem is that people don't like to think of themselves as being so predictable that a "robot" can guess what they are looking for. Whatever the reasons, the public has been staunchly resistant to AI in customer service, particularly on the phone, so be advised before putting your expensive and well-trained algorithms on your customer service desk. They'll probably be much happier hanging out with marketing or the web developers in the back office.
Do you want more information and insight for today's forward-thinking CIO? Get CIO Brief delivered to your inbox.